Some of the special terms mainly involved are: prediction function, cost function, error, gradient Descent, convergence, regularization, backpropagation, etc. executive email list , interested students can pay attention. 02 Supervised Learning Supervised learning is the machine learning task of inferring a function from labeled training data. The main feature is that the executive email list training data is labeled. For example, image recognition: when you input a picture of a cat, you tell the machine that it is a cat; when you input a picture of a dog, you tell the machine that it is a dog; so train.
In the test, when you input a photo that the machine has not seen before, the machine can identify whether the picture is a cat or a dog. The training data o executive email list f supervised learning is labeled, that is, when the input is to tell the machine what it is, the input is the data of the given label, so that the machine can automatically find out the relationship between the input and the output. In fact, most of the artificial intelligence we see executive email list now is supervised learning. Supervised learning advantages and disadvantages :
Advantages: The difficulty of the algorithm is low, and the model is easy to train; Disadvantages: Need to manually label a large amount of training data, thus giving executive email list birth to the positions of data labelers and data trainers. 03 Linear Regression The example of predicting housing prices we cited earlier is a linear regression problem. We need to find a line to executive email list fit these test data to minimize the error. If we want to minimize the error, that is, to minimize the sum of the distances from each test point to the straight line. (